Review URL: http://codereview.chromium.org/2095019

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@4690 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

Already reviewed at: http://codereview.chromium.org/2078022/show

TBR=vegorov@chromium.org
Review URL: http://codereview.chromium.org/2137008

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@4689 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

TBR=ager@chromium.org
Review URL: http://codereview.chromium.org/2071020

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@4688 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

the issue of how the pointee is found and how the pointer
is encoded are separated out.  This will make it simpler to
support various pointers from and to code in the future.
Review URL: http://codereview.chromium.org/2069013

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@4687 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

Review URL: http://codereview.chromium.org/2078022

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@4686 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

-- layout change for FixedArrays: length is stored as a smis (initial patch by Kevin Millikin)
-- layout change for SharedFunctionInfo: integer fields are stored as smi on arm, ia32 and rearranged on x64.
-- layout change for String: meaning of LSB bit is fliped (1 now means hash not computed); on x64 padding is added.
-- layout of maps is _not_ changed. Map space is currently iterated in a special way.
- Cardmarking write barrier. New barrier handles large objects and normal objects in a similar fashion (no more additional space for pointer tracking is required, no conditional branches in WB code).

Review URL: http://codereview.chromium.org/2101002

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@4685 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

Review URL: http://codereview.chromium.org/2076010

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@4684 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

Make sure that there is always room for RelocInfoWriter::kMaxSize
bytes in the buffer before writing to it. We only ensured space for 8
bytes, but relocation information can be up to 16 bytes in x64.

Review URL: http://codereview.chromium.org/2100010

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@4683 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

The Mozilla test ecma/Date/15.9.5.10-2 have been timing out a couple of times on the ARM simulator in debug mode.
Review URL: http://codereview.chromium.org/2070018

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@4682 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

Previously the receier was passed in both r0 and on the stack for a load IC. With this change the receiver is in r0 only.
Review URL: http://codereview.chromium.org/2119007

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@4681 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

The full compiler will now be used for all code compiler when debugging is active. As the code generated by the full compiler is much simpler it will be easier to make debugging work better when using that code.

To ensure that all code debugged is from the full compiler all functions will have to be recompiled when starting debugging. Initialing debugging already turns off the code cache.
Review URL: http://codereview.chromium.org/2120009

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@4680 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

The port to ARM of a tail call wrongly kept the scheme of pushing below the return address.

TBR=antonm@chromium.org
Review URL: http://codereview.chromium.org/2091013

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@4679 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

Using rsp or r12 as the "base" register of the ModR/M byte forces a SIB byte,
even with no index register. Some operations can avoid this by using another,
equivalent, encoding that swaps the meaning of the base and register parts.

Review URL: http://codereview.chromium.org/2075010

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@4678 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

ry (coldredlemur@gmail.com).  See http://codereview.chromium.org/2092007


git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@4677 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

This makes the full compiler handle all constructs on ia32. However the syntax checker for the full compiler is still the same so for both normal operation and with the flag --always-full-compiler the coverage of the full compiler will be the same.

This is on preparation for improving the debugger break point experience where the plan is to only use code from full code generator when debugging JavaScript.

Runs all tests on all three platforms in release and debug mode. The tests also run with both the following flags to the test runner

  --special-command="@ --nofull-compiler"
  --special-command="@ --always-full-compiler"

The changes to the x64 and ARM architectures are mainly structural due to the change to EmitVariableAssignment to handle initialization of const variables.
Review URL: http://codereview.chromium.org/1989012

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@4676 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

Collect cumulative (--print-cumulative-gc-stat) and per collection (--trace-gc-nvp) GC statistics and output it in a machine-readable name=value format.

Review URL: http://codereview.chromium.org/2132005

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@4675 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

Review URL: http://codereview.chromium.org/2129010

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@4674 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

As several pages can run in a single V8 instance, it is possible to
have functions from different security contexts intermixed in a single
CPU profile.  To avoid exposing function names from one page to
another, filtering is introduced.

The basic idea is that instead of capturing return addresses from
stack, we're now capturing JSFunction addresses (as we anyway work
only with JS stack frames.)  Each JSFunction can reach out for
context's security token. When providing a profile to a page, the
profile is filtered using the security token of caller page. Any
functions with different security tokens are filtered out (yes, we
only do fast path check for now) and their ticks are attributed to
their parents.

Review URL: http://codereview.chromium.org/2083005

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@4673 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

Review URL: http://codereview.chromium.org/2136012

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@4672 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

Review URL: http://codereview.chromium.org/2103007

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@4671 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

Also simplify the assertion code for type information in the macro assembler.

Review URL: http://codereview.chromium.org/2130003/show



git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@4670 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

TBR=ager@chromium.org
Review URL: http://codereview.chromium.org/2117010

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@4669 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

The calling convention for keyed store IC on ARM is changed to receive the arguments value, key and receiver in registers r0, r1 and r2 instead of on the stack. When calling keyed store IC with a virtual frame the arguments are passed through the virtual frame and consumed by the call.

Changed the register usage in the IC code to postpone spilling the registers holding value, key and receiver to the stack until making  a call into the runtime system.

Runs all the tests with: 

  --special-command="@ --nofull-compiler"
  --special-command="@ --always-full-compiler"
  --special-command="@ --noenable-vfp3"
Review URL: http://codereview.chromium.org/2116003

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@4668 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

When ldrd is not available two ldr instructions are generated. This fixes these in the case where the register used in the memory operand is the same as the first register in the register pair receiving the values.

All tests now run on ARM with the flag --special-command="@ --noenable-vfp3". Running without VFP3 support in the simulator causes more ldrd instructions to be used, and the default build configuration does not utilize ldrd, but generated tow ldr instructions.
Review URL: http://codereview.chromium.org/2078013

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@4667 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

Chromium build.

v8.gyp no longer sets any V8_TARGET_ARCH_* macro on the Mac. Instead, the
proper V8_TARGET_ARCH_* macro will be set by src/globals.h in the same way as
the V8_HOST_ARCH_* macro when it detects that no target macro is currently
defined. The Mac build will attempt to compile all ia32 and x86_64 .cc files.
#ifdef guards in each of these target-specific source files prevent their
compilation when the associated target is not selected. For completeness,
these #ifdef guards are also provided for the arm and mips .cc files.

BUG=706
TEST=x86_64 Mac GYP/Xcode-based Chromium build (still depends on other changes)
Review URL: http://codereview.chromium.org/2133003

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@4666 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

dtoa.c:69 states: #define Long int on machines with 32-bit ints and 64-bit
longs. This change was made in Chromium's version of dtoa in r21901. I don't
know why this apparently hasn't caused a problem in v8 until now. I was
unable to build a snapshot on x86_64 Mac without this change.

BUG=706
TEST=x86_64 Mac GYP/Xcode-based Chromium build (still depends on other changes)


git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@4665 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

r3995 (http://code.google.com/p/v8/source/detail?r=3995) introduce performance
regression for the case when arrat size is small (think new Array(4)).

It turns out that in those cases rep stos is slower than plain loop (apprently
due to ecx increment, but I didn't check this hypothesis.)  The next thing
to try could be direct jump into right place of long sequence of stos'es.

Review URL: http://codereview.chromium.org/2082006

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@4664 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

under one flag.  Also other cosmetic changes to peephole
optimization.
Review URL: http://codereview.chromium.org/2104006

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@4663 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

http://codereview.chromium.org/2004006 for Subrato De.


git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@4662 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

Review URL: http://codereview.chromium.org/2108006

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@4659 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

benchmark suite thus forming a new candidate benchmark revision:

  * Remove dead code from RayTrace benchmark.
  * Avoid repeated number-to-string conversion in Splay.

Review URL: http://codereview.chromium.org/2086005

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@4658 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

Review URL: http://codereview.chromium.org/2132003

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@4657 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

Review URL: http://codereview.chromium.org/2050007

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@4656 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

V8_INTERPRETED_REGEXP (as per Issue 1635001). 

BUG=none 
TEST=Ran v8 tests, as well as against current Chromium browser.

Review URL: http://codereview.chromium.org/2078005


git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@4655 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

Review URL: http://codereview.chromium.org/2094005

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@4654 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

Refactored code so global/non-global regexps are handled in separate functions.
Inlined ApplyReplaceFunction at its only call point.

Review URL: http://codereview.chromium.org/1994019

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@4653 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

TBR=sgjesse@chromium.org

Review URL: http://codereview.chromium.org/2033009

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@4652 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

Review URL: http://codereview.chromium.org/2015017

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@4651 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

If --nolazy is on, we compile ArraySort which requires these caches
to be set up.

Review URL: http://codereview.chromium.org/2025006

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@4650 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

Bailout to JS Array builtins if array's prototype is different from
Array.prototype.  Otherwise there might be inherited elements coming
from this prototype.

Review URL: http://codereview.chromium.org/2037008

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@4649 ce2b1a6d-e550-0410-aec6-3dcde31c8c00